Simulating extreme cold spells in France with empirical importance sampling

Extreme winter cold spells in Europe have huge societal impacts. Being able to simulate worst case scenarios of such events for present and future climates is hence crucial for adaptation. Rare event algorithms have been applied to simulate extreme heatwaves. They have emphasized the role of the atmospheric circulation in such extremes. The goal of this study is to test such algorithms to extreme cold spells.
We focus on cold spells that occur in France since 1950. The analysis is based on the ERA5 reanalysis. We select cold events that have occurred for different time scales (10 days, 1 month, 3 months). We identify record shattering cold events for time scales of 1 and 3 months (in 1956 and 1963). We find that, although the frequency of extreme cold spells decreases with time, their intensity is stationary.
We applied a stochastic weather generator approach with importance sampling, to simulate the worst cold spells that could occur every year since 1950, with lengths of 1 month and 3 months. We hence simulated ensembles of worst winter cold spells that are consistent with observations. Those worst cases are slightly colder than the record shattering events, and do not yield the trend that is observed on the mean temperature. The atmospheric circulation that prevails during those events is analyzed and compared to the observed circulation during the record breaking events.